The determination of NT-proBNP levels is key for the monitoring of patients with diagnosed heart failure and it is a routine measurement typically performed at health care centers, which would benefit from decentralized alternatives. Here we investigate the quantitative evaluation of a well-established NT-proBNP test using a standard mobile phone (Nokia 6720) as measuring platform rather than a dedicated instrument. A Java ME software developed for this application controls the illumination and imaging of the proBNP test under defined time intervals, which enables the composition of multi-exposure sets that are processed as high dynamic range (HDR) images for contrast enhancement. The results show that HDR processing significantly increases the sensitivity and resolution of the technique achieving a performance within the diagnostics range. These results demonstrate the feasibility to exploit a ubiquitous device to decentralize the evaluation of a routine test and identify key processing alternatives to bring the performance of such systems within the diagnostics range. © 2010 Elsevier B.V. All rights reserved.